J. Embryol. exp. Morph. 73, 221-232, 1983
Printed in Great Britain © The Company of Biologists Limited 1983
221
Determination of the curvature of epithelial cell mass
by mesenchyme in branching morphogenesis of mouse
salivary gland
By HIROYUKI NOGAWA 1
From the Biological Laboratory, College of Arts and Sciences,
Chiba University
SUMMARY
Mouse submaxillary epithelium undergoes branching morphogenesis with increase in the
curvature of its surface in vivo. Recombination experiments in vitro of the epithelium and
mesenchyme between 13- and 14-day rudiments showed (1) that the 14-day mesenchyme more
actively induced the epithelium to branch than the 13-day mesenchyme, (2) that the 14-day
mesenchyme could produce clefts on smaller epithelial lobes than the 13-day mesenchyme, (3)
that the 14-day mesenchyme produced lobes with similar diameter to lobes of the 14-day intact
rudiment, and (4) that the lobular morphology of assembled 14-day lobes became obscure in
recombinates with the 13-day mesenchyme while it was well maintained in recombinates with
the 14-day mesenchyme. From these results it is concluded that the mesenchyme determines
the curvature of epithelial surface, and that clefts are formed on the epithelial surface as a
result of increase in the epithelial curvature.
INTRODUCTION
Mechanisms of branching morphogenesis of mouse salivary gland have been
investigated by many workers since introduction of in vitro culture of this organ
by Borghese (1950). Spooner & Wessells (1972) and Spooner (1973) presented a
model of salivary branching morphogenesis in which clefts were formed by
changes of the shape of epithelial cells which were caused by shortening or
lengthening of epithelial microfilament bundles, and deeper clefts were stabilized
by collagen fibrils at the epitheliomesenchymal interface. Thereafter, this model
was modified by Banerjee, Cohn & Bernfield (1977), taking account of basal
lamina which consisted of proteoglycans and had a role of maintaining epithelial
morphology. Spooner & Faubion (1980) and Thompson & Spooner (1982) reconfirmed the role of collagen and proteoglycans as stabilizers of the epithelial morphology using inhibitors of synthesis of these substances respectively.
Studies on epitheliomesenchymal interactions in mouse salivary morphogenesis have shown that only a few kinds of other mesenchymes than the
1
Author's address: Biological Laboratory, College of Arts and Sciences, Chiba University,
Yayoi-cho, Chiba 260, Japan.
EMB73
222
H. NOGAWA
salivary mesenchyme can support branching of the salivary epithelium (Grobstein, 1953; Sherman, 1960; Cunha, 1972; Lawson, 1974), and that the salivary
mesenchyme can instructively influence mammary epithelium to branch in
salivary-like fashion (Kratochwil, 1969; Sakakura, Nishizuka & Dawe, 1976).
We have introduced the elongating and branching morphogenesis of quail salivary glands to study tissue interactions in the epithelial morphogenesis (Nogawa,
1978 & 1981; Nogawa & Mizuno, 1981). Recombination experiments of the
epithelium and mesenchyme between elongating-type (quail anterior submaxillary) and branching-type (quail anterior lingual and mouse submaxillary) salivary glands showed that the elongating or branching morphogenesis of quail
salivary epithelium was controlled by the associated mesenchyme. Yet, it
remains unclear how the salivary mesenchyme participates in controlling the
epithelial morphogenesis.
Comparative observations of the branching and elongating morphogenesis in
vivo let us notice that the curvature of branching-type epithelium became larger
with progress of the development (Fig. 1), while that of elongating-type
epithelium was nearly constant during elongation (see Figs 2 to 4 in Nogawa,
1981). This may suggest that 'branching' can be regarded as increase in the
epithelial curvature. In the present study, branching morphogenesis of mouse
submaxillary gland is analysed with recombination experiments in vitro of the
epithelium and mesenchyme between two submaxillary rudiments whose
epithelial curvature is different, and a new model of branching morphogenesis
is presented from a viewpoint of the epithelial curvature determined by the
associated mesenchyme.
MATERIALS AND METHODS
Isolation of submaxillary rudiments
ICR mice were mated during the night, and the morning of discovery of the
vaginal plug was counted as day 0. Submaxillary rudiments with sublingual
rudiments were isolated from 13-, 14- and 15-day foetuses according to the
procedure of Borghese (1950), and accompanied sublingual rudiments were
removed.
Separation of epithelium and mesenchyme
Isolated 13- and 14-day submaxillary rudiments were exposed to 0-25 % trypsin (1:250) in Ca 2+ -, Mg2+-free Hanks' balanced salt solution (GIBCO
Laboratories) at 4 °C for 15 min. Since the distal part of the rudiments consisted
of only mesenchymal component, the submaxillary mesenchyme was taken up
from the distal part. The submaxillary epithelium was cleared of mesenchymal
cells with very fine forceps, and a whole part of the 13-day epithelium or lobes
of the 13- and 14-day epithelium were used in recombination experiments.
Curvature of salivary epithelium determined by mesenchyme 223
Separated epithelia and mesenchymes were rinsed twice in a mixture of Hanks'
balanced salt solution (HBSS) and horse serum (1:1), and stored in HBSS with
20% horse serum at room temperature. Before being recombined with the
mesenchyme, 13- and 14-day epithelial lobes were measured, and 13-day lobes
with diameter 135-150/im and 14-day lobes with diameter 95-110 pan were used.
Organ culture
Explants were cultivated at 37-5 °C according to the method of Wolff & Haffen
(1952). The medium was composed of medium 199 (GIBCO Laboratories) with
20 % horse serum (Flow Laboratories), 0-5 % agar (DIFCO Laboratories) and
penicillin G potassium (100 units/ml). Three pieces of 13- or 14-day submaxillary mesenchyme were assembled on the semisolid medium, and submaxillary
epithelia were placed in the centre of the mesenchymal mass.
Measurement
Recombinates were photographed 20 or 44 h after cultivation. All the clefts
formed on the epithelial surface in each recombinate were classified as deep
clefts (deeper than 50[xm) and shallow clefts, and the value of cleft formation
(Vc) was given to each recombinate by counting one deep cleft as 1 and one
shallow cleft as 0-5.
In order to compare the size of epithelial lobes with shallow clefts and those
without clefts, the area of epithelium which was traced on the thick paper was
weighed, and diameter of the lobe was calculated as a circle from the weight.
RESULTS
Increase of epithelial curvature in normal development
A submaxillary epithelium of the 13-day rudiment consists of three to five
spherical lobes and a cylindrical stalk (Fig. 1A). Diameter of the lobes is 120180/im, and that of the stalk is 80-100fim. On the 14th day, the epithelium
undergoes branching and the number of lobes becomes 30-40 (Fig. IB).
Diameter of the lobes decreases (70-130 /im), and that of the stalk also decreases
(40-50jum). On the 15th day, the number of lobes reaches 150-200 (Fig. 1C).
Diameter of the lobes continues to decrease (40-100/im), but that of the stalk
is constant. These observations show that the curvature of lobular epithelium
increases with developmental stages, and that lobes with diameter over 130 fim,
which are present in the 13-day rudiment, are absent in the 14-day rudiment.
Morphogenesis of 13-day epithelium in recombinates
In intact rudiments, the 14-day mesenchyme surrounded smaller lobes than
the 13-day mesenchyme. When a whole 13-day epithelium was cultured in
recombination with the 14-day mesenchyme for 20 h, the epithelium formed
224
H. NOGAWA
Fig. 1. Submaxillary rudiments of 13- (A), 14- (B) and 15-day foetuses (C).
Bar = 500jum.
smaller and more lobes than the epithelium did when cultured in recombination
with the 13-day mesenchyme (Figs 2A and B).
In order to simplify the experimental system, one of 13-day epithelial lobes
with diameter 135-150 jum was recombined with the 13- or 14-day mesenchyme
and cultured (Fig. 3A, Table 1). The value of cleft formation (Vc) of the lobe
recombined with the 13-day mesenchyme was 1 • 1 ± 1 • 1 at 20 h of cultivation, and
the epithelium formed shallow clefts in two thirds of the recombinates (Fig. 3B).
In contrast, the Vc of the lobe recombined with the 14-day mesenchyme was
3-4 ± 1-1 at 20 h of cultivation, and the epithelium cleaved deeply to form new
lobes in most of the recombinates (Fig. 3C). Diameter of the newly formed lobes
was 110 ± 20jum (n = 54, min. 50/im and max. 160 jum) and was similar to that
of one lobe of the 14-day intact rudiment. A question then arose as to how the
morphogenesis of a smaller epithelial fragment of the 13-day lobe proceeded
Curvature of salivary epithelium determined by mesenchyme 225
Fig. 2. Recombinates of a whole 13-day epithelium with the 13-day mesenchyme
(A) and with the 14-day mesenchyme (B) at 20 h of cultivation.
Fig. 3. Recombinates of one 13-day lobe with the mesenchymes. The lobe (arrow)
at the beginning of cultivation (A). The lobe recombined with the 13-day mesenchyme (B) forms only a shallow cleft at 20 h of cultivation, while the lobe recombined
with the 14-day mesenchyme (C) forms deeper clefts and consists of several lobes.
Fig. 4. Recombinates of the one-third 13-day lobe with the 13-day mesenchyme (A)
and with the 14-day mesenchyme (B) at 20 h of cultivation. Two shallow clefts are
formed in B.
Bar = 200Jum.
when cultured in recombination with the mesenchymes. One third of the one
13-day lobe was cut off and cultured in recombination with the 13- or 14-day
mesenchyme (Table 1). In most of the recombinates with the 13-day mesenchyme
226
H. NOGAWA
Table 1. Cleft formation of the 13-day epithlium in recombinates at 20 h of
cultivation
No. of recombinates
(diameter of epithelium, jum)*
A.
Epithelial
lobe
Age of
mesenchyme
(day)
with
deep and
shallow
clefts
with
only
shallow
clefts
one
13
4
17
(240 ± 30)
4
(230 ± 20)
3
(180 ± 10)
11
(170 ± 20)
1/3
14
20
13
0
14
1
total
Value of *
cleft
formation
(Vc)
26
1-1 zh 1 1
0
24
3-4 ±1-1
16
(170 ± 20)
9
(160 ± 10)
19
0-1 ±0-2
21
0-7 + 0-9
without
clefts
5
(240 ± 20)
*Mean ±s. D.
the epithelial fragment remained round at 20 h of cultivation (Fig. 4A), but
underwent branching at 44 h. In contrast, half of the recombinates with the
14-day mesenchyme formed shallow clefts at 20h (Fig. 4B), while the rest half
remained round.
Recombinates with the 13-day mesenchyme and those with the 14-day mesenchyme were compared on diameter of lobes which were forming shallow clefts
(Table 1). When the 13-day mesenchyme was associated, the lobe with diameter
ca. 240 jum 20 h after cultivation could form shallow clefts, but the lobe with
diameter ca. 170/xm could not. However, when the 14-day mesenchyme was
associated, even the lobe with diameter ca. 170//m 20 h after cultivation could
form shallow clefts.
Morphogenesis of 14-day epithelium in recombinates
One of the 14-day epithelial lobes with diameter 95-110 /im was recombined
with the 13- or 14-day mesenchyme and cultured (Fig. 5A, Table 2). In most of
the recombinates with the 13-day mesenchyme the lobe remained round at 20 h
of cultivation (Fig. 5B), but underwent branching at 44 h. In contrast, the lobe
formed shallow clefts at 20 h of cultivation in most of the recombinates with the
14-day mesenchyme (Fig. 5C), and the Vc was 1-0 ± 0-7, which was similar to the
Vc of one 13-day lobe with the 13-day mesenchyme. Diameter of the 14-day lobe
forming shallow clefts in recombinates with the 14-day mesenchyme was ca.
180/im, which was similar to that in recombinates of the one-third 13-day lobe
with the 14-day mesenchyme. These data indicated that the 14-day mesenchyme
could produce shallow clefts on the lobe with diameter ca. 180 /mi irrespective
of age of the epithelium.
Curvature of salivary epithelium determined by mesenchyme 227
Fig. 5. Recombinates of one 14-day lobe with the mesenchymes. The lobe (arrow)
at the beginning of cultivation (A). The lobe recombined with the 13-day mesenchyme (B) remains round at 20 h of cultivation, while the lobe recombined with the
14-day mesenchyme (C) forms three shallow clefts.
Fig. 6. Recombinates of an assembly of two 14-day lobes with the mesenchymes.
The assembled two lobes (arrows) at the beginning of cultivation (A). The two
original lobes fuse into one larger lobe without clefts in recombinate with the 13-day
mesenchyme at 20 h of cultivation (B), while the fusing lobes have a few clefts besides
two deep clefts in recombinate with the 14-day mesenchyme (C).
Fig. 7. Recombinates of an assembly of three 14-day lobes with the mesenchymes.
The assembled three lobes (arrows) at the beginning of cultivation (A). The three
original lobes fuse into one larger lobe with three clefts in recombinate with the
13-day mesenchyme at 20 h of cultivation (B), while the fusing lobes have a few clefts
besides three deep clefts in recombinate with the 14-day mesenchyme (C).
Bar = 200 jum.
228
H. NOGAWA
Table 2. Cleft formation of the 14-day epithlium in recombinates at 20 h of
cultivation
No. of recombinates
(diameter of epithelium, /mi)*
A
r
Epithelial
lobe
one
with
Age of
deep and
mesenchyme shallow
(day)
clefts
13
0
14
1
two
13
5
three
14
13
13
8
14
16
with
only
shallow
clefts
5
(190110)
19
(180110)
6
(230120)
1
5
(270120)
0
total
Value of*
cleft
formation
(Vc)
15
(170110)
4
(180120)
3
(250130)
0
0
20
0-210-4
24
1-010-7
14
1-210-9
14
13
2-7 ±0-7
2-210-9
0
16
4-911-2
without
clefts
* Mean 1 s.D.
Since the 13-day mesenchyme could not have a morphogenetic effect on one
14-day lobe 20 h after cultivation which was smaller than one 13-day intact lobe at
the beginning of cultivation, it was investigated whether the 13-day mesenchyme
could influence the morphogenesis of assembled 14-day lobes (Figs 6A and 7A,
Table 2). When two lobes were assembled and cultured in recombination with the
13-day mesenchyme, the lobes fused together in all the cases. At 20 h of cultivation, the fusing lobes rarely had more than two shallow clefts and a few of them
became one larger lobe without clefts (Fig. 6B). When an assembly of three lobes
was cultured in recombination with the 13-day mesenchyme, the fusing lobes
mostly had only three clefts at 20 h of cultivation which consisted of a mixture of
shallow and deep clefts (Fig. 7B). However, when an assembly of two or three
lobes was cultured in recombination with the 14-day mesenchyme, it formed new
clefts in addition to two or three deep clefts at 20 h of cultivation (Figs. 6C and 7C).
These results suggested that since the 13-day mesenchyme could not maintain the
morphology of individual 14-day lobes differently from the 14-day mesenchyme,
the boundaries between the lobes disappeared after fusion or became shallow
clefts in recombinates with the 13-day mesenchyme while they developed into
deep clefts in recombinates with the 14-day mesenchyme.
DISCUSSION
In the previous study it was demonstrated that mouse submaxillary mesenchyme induced the elongating-type quail salivary epithelium to branch (Nogawa
Curvature of salivary epithelium determined by mesenchyme 229
& Mizuno, 1981), but the nature of the instructive influence of the mesenchyme
remained unclear. The present study demonstrated that each lobe of mouse
submaxillary epithelium became smaller with progress of the development in
vivo. From this fact it was expected that making the curvature of lobes larger
resulted in cleaving lobes. Then, recombination experiments in vitro of the
epithelium and mesenchyme between 13- and 14-day rudiments showed that the
14-day mesenchyme more actively induced the epithelium to branch than the
13-day mesenchyme. The 14-day mesenchyme, however, did not more actively
stimulate the epithelial growth since there was no difference in diameter between
the lobes forming shallow clefts in recombinates with the 14-day mesenchyme
and the round lobes without clefts in recombinates with the 13-day mesenchyme.
Although Alescio & Di Michele (1968) reported that in the development of
mouse embryonic lung the branching activity of epithelium correlated to the
growth rate of epithelium, it is not appropriate to this case. The higher branchinducing activity of the 14-day mesenchyme seems to originate from the fact that
the mesenchyme has the ability to make the curvature of epithelial surface as
large as that of the intact lobe of the mesenchyme-dependent age: in the lobe
with diameter ca. 170/im, the 14-day mesenchyme could produce shallow clefts
while the 13-day mesenchyme could not, and in the lobe with diameter ca.
240 jum, the 14-day mesenchyme produced several lobes as small as the 14-day
intact lobes while the 13-day mesenchyme produced only shallow clefts. Probably, clefts may be formed on the epithelial surface as a result of that the mesenchyme increases the epithelial curvature. This supposition is supported by the
results in the further recombination experiments that the lobular morphology of
an assembly of two or three 14-day lobes was maintained by two or three deep
clefts in recombinates with the 14-day mesenchyme while it became obscure in
recombinates with the 13-day mesenchyme.
The present study strongly suggests that submaxillary mesenchyme causes the
epithelium to branch by determining the curvature of epithelial surface, but
actually the mesenchyme will determine the curvature of the basal lamina.
Banerjee et al. (1977) reported that mouse submaxillary epithelium came to
project processes when the basal lamina was torn off from the epithelial surface,
and that such epithelium became round with disappearance of clefts when cultured in recombination with the mesenchyme. The mesenchyme may not be able
to affect the epithelial surface whose curvature is irregular with many processes,
and to smooth the epithelial surface may be the most important role of the basal
lamina in branching morphogenesis.
By what mechanisms will submaxillary mesenchyme determine the curvature
of epithelial surface? Reports on contact guidance by Dunn & Heath (1976) and
Fisher & Tickle (1981) may suggest a possible mechanism: flbroblastic cells
cannot hold firmly enough to a substratum whose curvature is larger than a
critical curvature, and their microfilament bundles play an important role in this
cell behaviour. Submaxillary mesenchymal cells may have stage-dependent
230
H. NOGAWA
increasing critical curvature, and they may be able to affect the epithelial morphogenesis only when they can hold the epithelial surface. Spooner & Wessells
(1972) and Spooner (1973) postulated that branching morphogenesis was caused
by changes of epithelial microfilament bundles, but a possibility remains still that
mesenchymal microfilament bundles play an important role in determining the
curvature of epithelial surface.
It was also observed in the present study that several shallow clefts appeared
on the epithelial surface at a short interval in early 13-day mouse submaxillary
rudiments before branching (Fig. 8). From this finding and the results in the
recombination experiments a new model of branching morphogenesis is
proposed (Fig. 9). The curvature-increasing model, in which the branching morphogenesis is regarded not as formation of clefts at particular points of the
epithelial surface but as increase in the curvature of all the epithelial surface,
seems to explain well the dynamic aspect of branching morphogenesis. In normal
development of mouse submaxillary rudiments, the initiation of branching will
be ensured by the increase in the mesenchyme-determining curvature with
developmental stages. Then, how will the epithelial morphogenesis take place
when the mesenchyme-determining curvature is constant during development?
The epithelium will form a cylindrical stalk without branching, since quail anterior submaxillary rudiments elongate keeping the epithelial curvature constant
Fig. 8. A submaxillary rudiment of the early 13-day foetus. Several shallow clefts
(arrows) are formed at a short interval. Bar = 50/im.
Curvature of salivary epithelium determined by mesenchyme 231
c
g
Fig. 9. Curvature-increasing model of branching morphogenesis. A: when the
radius of epithelial curvature (re) is smaller than that of mesenchyme-determining
curvature (rm), the lobe remains round. Thereafter, re increases by epithelial cell
proliferation, and rm decreases with developmental stages. B: when re becomes
larger than r m , many arcs with rm appear on the epithelial surface under the influence
of surrounding mesenchyme, and consequently many shallow clefts are formed. B':
when the mesenchyme influences also shallow clefts to take the curvature of r m ,
some clefts disappear and others deepen. C: finally, spherical lobes with rm are
formed, and the number of lobes formed is determined by a ratio of total volume of
epithelium to the volume of one lobe with rm . The curvature of epithelial surface is
constant during stages B and C.
in vivo (Nogawa, 1981). If this consideration is added to the curvature-increasing
model, the model will design various patterns of branching morphogenesis of
glandular organs.
We have grasped in the present study the branching morphogenesis of mouse
submaxillary gland as increase in the curvature of epithelial surface under the
control of mesenchyme. Recently, Yang, Larson & Nandi (1982) succeeded in
making mouse submaxillary epithelial cells grow three dimensionally in
mesenchyme-free condition by using collagen gel, but the morphology of the
cell mass was quite different from the lobular morphology observed both in the
intact rudiment and in the recombinate. We will elucidate the mechanisms of
mesenchymal control over the curvature of salivary epithelium in future studies.
The author wishes to express his deep gratitude to Prof. Takeo Mizuno and Dr. Shigeo
Takeuchi of the University of Tokyo and staffs of the Chiba University for their valuable
advice and encouragement during the course of this work.
REFERENCES
T. & Di MICHELE, M. (1968). Relationship of epithelial growth to mitotic rate in
mouse embryonic lung developing in vitro. J. Embryol. exp. Morph. 19, 227-237.
BANERJEE, S. D., COHN, R. H. & BERNFIELD, M. R. (1977). Basal lamina of embryonic
salivary epithelia: production by the epithelium and role in maintaining lobular morphology. J. Cell Biol. 73, 445-463.
BORGHESE, E. (1950). The development in vitro of the submandibular and sublingual glands
of Mus musculus. J. Anat., Lond. 84, 287-302.
CUNHA, G. R. (1972). Support of normal salivary gland morphogenesis by mesenchyme
derived from accessory sexual glands of embryonic mice. Anat. Rec. 173, 205-212.
ALESCIO,
232
H. NOGAWA
G. A. & HEATH, J. P. (1976). A new hypothesis of contact guidance in tissue cells. Expl
Cell Res. 101,1-14.
FISHER, P. E. & TICKLE, C. (1981). Differences in alignment of normal and transformed cells
on glass fibres. Expl Cell Res. 131, 407-410.
GROBSTEIN, C. (1953). Epithelio-mesenchymal specificity in the morphogenesis of mouse submandibular rudiments in vitro. J. exp. Zool. 124, 383-414.
KRATOCHWIL, K. (1969). Organ specificity in mesenchymal induction demonstrated in the
embryonic development of the mammary gland of the mouse. Devi Biol. 20, 46-71.
LAWSON, K. A. (1974). Mesenchyme specificity in rodent salivary gland development: the
response of salivary epithelium to lung mesenchyme in vitro. J. Embryol. exp. Morph. 32,
469-493.
NOGAWA,H. (1978). The development of the salivary glands in the Japanese quail. 7. Fac. Sci.,
Univ. Tokyo IV14, 95-103.
NOGAWA, H. (1981). Analysis of elongating morphogenesis of quail anterior submaxillary
gland: absence of localized cell proliferation. /. Embryol. exp. Morph. 62, 229-239.
NOGAWA, H. & MIZUNO, T. (1981). Mesenchymal control over elongating and branching
morphogenesis in salivary gland development. J. Embryol. exp. Morph. 66, 209-221.
SAKAKURA, T., NISHIZUKA, Y. & DAWE, C. J. (1976). Mesenchyme-dependent morphogenesis
and epithelium-specific cytodifferentiation in mouse mammary gland. Science. 194,
1439-1441.
SHERMAN, J. E. (1960). Description and experimental analysis of chick sub-mandibular gland
morphogenesis. Trans. Wis. Acad. Sci. 49, 171-189.
SPOONER, B. S. (1973). Microfilaments, cell shape changes, and morphogenesis of salivary
epithelium. Amer. Zool. 13, 1007-1022.
SPOONER, B. S. & FAUBION, J. M. (1980). Collagen involvement in branching morphogenesis
of embryonic lung and salivary gland. Devi Biol. 77, 84-102.
SPOONER, B. S. & WESSELLS, N. K. (1972). An analysis of salivary gland morphogenesis: role
of cytoplasmic microfilaments and microtubules. Devi Biol. 27, 38-54.
THOMPSON, H. A. & SPOONER, B. S. (1982). Inhibition of branching morphogenesis and
alteration of glycosaminoglycan biosynthesis in salivary glands treated with /3-D-xyloside.
Devi Biol. 89, 417-424.
WOLFF, ET. & HAFFEN, K. (1952). Sur une mdthode de culture d'organes embryonnaires "in
vitro". Tex. Rep. Biol. Med. 10, 463-472.
YANG, J., LARSON, L. & NANDI, S. (1982). Three-dimensional growth and morphogenesis of
mouse submandibular epithelial cells in serum-free primary culture. Expl Cell Res. 137,
481-485.
DUNN,
{Accepted 26 August 1982)